Order of U+103A (Asat) and U+1037 (Aukmyit) in Unicode 5.1 encoding input

[GoogleCodeExporter]

#Is this a feature request? If so, describe your feature here:

#Is this a bug report? If so, describe the bug in general:
In Unicode 5.1 encoding input, U+103A U+1037 should be normalized to U+1037 
U+103A in all rhymes where aukmyit followed killed consonant.

#Describe steps necessary to reproduce this bug:
Type romanisation "thant"

#What is the expected output? What do you see instead?
Expected output - U+101E U+1014 U+1037 U+103A
Wrong output - U+101E U+1014 U+103A U+1037

#What version of WaitZar do you have? What version of Windows?
#  (Click on the WZ icon and choose "Help" to see the version number)
libwaitzar_1.0.1, ibus-waitzar_0.0.1, fedora 14

Original issue reported on code.google.com by tmh...@gmail.com on 27 Nov 2010 at 3:46

• Blocking: #131

Attachments:

• wrong.png
• correct.png

[GoogleCodeExporter]

Oh, nice catch. Yes, you are correct; the two should be reversed. 

Interestingly, Padauk displays the (correct) rendering correctly, but Myanmar3 
does not. (See attached). Perhaps this is just in Windows; I'll have to try it 
later in Ubuntu.

The 1.9 release of WZ will (probably) switch to storing Unicode internally (now 
it still converts from Zawgyi). I will update ibus-waitzar to use this too, and 
we can fix all the remaining normalization errors. :)

Original comment by seth.h...@gmail.com on 27 Nov 2010 at 3:56

• Changed state: Confirmed

Attachments:

• pd_mm3.png

[GoogleCodeExporter]

Thanks for fast response. Myanmar3 haven't updated yet for this issue.

Sadly, I cannot even add in mywords.txt as it's converted from Zawgyi.

Waiting for 1.9 release and next release of ibus-waitzar.

Cheers

Original comment by tmh...@gmail.com on 27 Nov 2010 at 4:34

[GoogleCodeExporter]

I found an updated version of Myanmar3 that has the ordering correct:
  http://www.myanmarnlp.org.mm/images/stories/download/fonts.zip

I've attached the correct font, too. Tested in Open Office; it shows the text 
correctly.

We actually ran into this issue when translating the User's Guide. 

Original comment by seth.h...@gmail.com on 29 Nov 2010 at 1:09

Attachments:

• mm3.ttf

[GoogleCodeExporter]

Thanks for info :)

Original comment by tmh...@gmail.com on 30 Nov 2010 at 4:40